In this section I look into the overlaps between the phenomena of borrowing and codeswitching.
I present different ways various scholars use to distinguish them. Towards the end of this section I will outline the ways the phenomena are distinguished in this thesis.
According to Onysko (2007: 36), the phenomena of codeswitching and borrowing often overlap. Many scholars have spent pages and pages in order to distinguish between the two phenomena (e.g. Myers-Scotton 1993; Myers-Scotton 2006; Poplack 2017; Callahan 2004;
McClure 1998). One of the ways to distinguish between these two is Poplack’s (1993: 255) idea presented in section 2.3 that in codeswitching the word or phrase follows the morphological, syntactic and, in speech, phonological, rules of the language of its provenance. Lauttamus (1990: 12) is in line with Poplack as he mentions that the lack of inflection would make an utterance a codeswitch rather than borrowing, and vice versa: if the utterance is inflected according to the rules of the matrix language, or receptor language, it would rather be a borrowing. The arguments against this view were raised in section 2.3, and, since, the Finnish language requires inflection quite often, in this thesis codeswitches do sometimes follow the morphological rules of the matrix language, such as Finnish inflections. Hence, these are not valid ways to distinguish between the two phenomena in this thesis.
If single-word instances of foreign origin can be either codeswitches or loans, how to distinguish between them? Since one of the stages of assimilation of loan words, the citation loans (Sajavaara 1989: 97), have not been integrated into the recipient language, they might resemble codeswitches. Yet, Onysko (2007: 36–37) mentions a few ways in which codeswitches can be distinguished from borrowings. First, he mentions the notion made by Clyne (2003: 71), that while codeswitches can be either single-word or multi-word elements, borrowing is only limited to the former. Furthermore, borrowings belong to the lexicon of the
29
matrix language but codeswitches do not; the latter belong to the embedded language lexicon (Onysko 2007: 36).
Callahan (2004: 9) mentions that frequency is often used when trying to distinguish between borrowings and codeswitches. She explains, that words that are accepted in the matrix language lexicon will appear more frequently in texts. Thus, the reason is, in essence, the same as presented by Onysko (2007: 36): the loan words are a part of the matrix language lexicon, which makes them more available to all speakers of the matrix language, while codeswitches often require some level of bilingual competence. Hence, also availability of the utterance to a monolingual speaker (or writer) versus a bilingual speaker (or writer) can be used to distinguish between borrowings and codeswitches. Yet, as discussed in the previous section, flagging devices, such as translation, can be used to make the codeswitch utterances more available to a monolingual listener or reader.
Myers-Scotton (1993: 163), among others (see e.g. Seitamäki 2013; Callahan 2004) views loan words and codeswitches rather as a continuum than separate phenomena. Myers-Scotton (2006:
254) argues that many singly occurring codeswitches eventually become established loan words, especially if they are adopted by “trend-setters”. She utilizes Levelt’s (1989: 6) idea of mental lexicon, which is “the store of information about the words in one’s language”. She notes that, even though forming codeswitches and borrowings in one’s speech or text goes through the same production process, only the borrowings, or loan words, are part of the mental lexicon of the matrix language (Myers-Scotton 1993: 163).
This aforementioned difference in the status of codeswitches and loan words in regard to the lexicon of the matrix language, leads, according to Myers-Scotton (1993: 163, 176), to the notable differences in the frequency of the instances. Thus, like Callahan (2004: 9), she sees frequency as one way to distinguish between codeswitches and loan words. In fact, Myers-Scotton (1993: 176) states that the only criteria that holds in all cases when differentiating between the two phenomena are the absolute and relative frequencies of the instances.
On the other hand, Onysko (2007: 38) points out that with a limited corpus it is highly difficult to determine whether an instance that appears only once in the study material will continue to exist in the lexicon of the matrix language in question. This means that, by examining only a limited amount of text, it might be difficult to distinguish between codeswitches and loan words,
30
if frequency is the only criterion. McClure (1998: 131) notes the difficulties with frequency as well by pondering on how much frequency is enough for an instance being considered established (i.e. a loan word).
Even though using frequency as an indicator of codeswitches can be problematic, differences between the frequency of codeswitches and the frequency of loan words was still visible in my study. Infrequency of codeswitch-instances was a norm in my material, as most of the codeswitches analyzed were names of songs, movies and TV programs and hence did not appear in the texts but a few times. In contrast, many loan words appeared in the material several times. Similarly, other types of codeswitches, such as the example mentioned in section 2.3, bullshittinä, occurred only once. Yet, the frequency of the instances could not have been the deciding factor as some clear, old loan words, such as traktori (Häkkinen 2009: 1338) occurred likewise only once in my data, but it would be quite absurd to claim that traktori is a codeswitch due to infrequency. Moreover, as the data used in this study is rather small, one cannot draw overall conclusions on the frequency of the words in overall language use; in other contexts some of the instances found only once in this data could be highly frequent, and vice versa.
Thus, in this thesis, the deciding factor between loan words and codeswitches is the Finnish lexicon and whether the instances are accepted and established in it. Consequently, I utilized several etymological dictionaries (see section 3.2) and, similarly to McClure (1998: 132), a monolingual dictionary, more specifically the dictionary of Institute for the Languages of Finland, Kielitoimiston sanakirja (2018), to confirm whether an instance was part of the Finnish lexicon or not. If it was not, it was considered a codeswitch and if it was, it was considered a direct loan (if it met the other criteria set for direct loans in section 2.2).
In this part of the thesis I have introduced the background literature that discusses the phenomena I study: direct loans, hybrids and codeswitches. At the same time, I have defined the terminology I use in this study. Next, I will move on to presenting the material and methods of the study, after which I will introduce the results of the analysis.
31